基于MVDR的宽带水下被动声自导系统远程目标检测方法  

Detection of Long-distance Object Based on MVDR for Wideband Underwater Passive Acoustic Home Guide System

作　　者：游鸿[1] 黄建国[1] 

机构地区：[1]西北工业大学航海学院,陕西西安710072

出　　处：《兵工学报》2009年第2期160-164,共5页Acta Armamentarii

基　　金：国家自然科学基金资助项目(60572098)

摘　　要：远程目标检测是水下航行器被动声自导系统的重要技术问题。本文采用基于子带分解的宽带最小方差无畸变响应(MVDR)波束形成进行空间能量检测,通过在频域各子带进行自相关矩阵平滑有效改善检测稳定性。相比传统窄带及未进行自相关矩阵平滑的宽带波束形成方法,所提方法能显著提高系统低信噪比检测性能。仿真和湖试实验结果表明该方法有较好的工程应用前景。The detection of long distance object is a important technical area for passive home guide system in underwater vehicle.The detection method of long-distance object was proposed by minimum variance distortionless response（MVDR） beam forming,based on sub-band decomposition,to detect spatial energy distribution accurately.On the basis of the method,the detecting stability was improved by auto-correlation matrix smoothing in every sub-band.It uses auto-correlation matrix smoothing in every sub-band to improve the stability of detection.It can improve the detection performance under low SNR condition as compared with conventional narrowband or wideband beam forming methods without auto-correlation matrix smoothing.Computer simulation and authentication in lake show it is hopeful in engineering applications.

关 键 词：信息处理技术 弱信号检测 阵列信号处理 宽带信号 最小方差无畸变响应 

分 类 号：TN911.72[电子电信—通信与信息系统]